| | --- |
| | base_model: save_model/llama3-2k-non-ref-spa_2 |
| | tags: |
| | - alignment-handbook |
| | - generated_from_trainer |
| | datasets: |
| | - datasets/training-llama3-2k-non-ref-spa_3 |
| | model-index: |
| | - name: llama3-2k-non-ref-spa_3 |
| | results: [] |
| | --- |
| | |
| | <!-- This model card has been generated automatically according to the information the Trainer had access to. You |
| | should probably proofread and complete it, then remove this comment. --> |
| |
|
| | # llama3-2k-non-ref-spa_3 |
| | |
| | This model is a fine-tuned version of [save_model/llama3-2k-non-ref-spa_2](https://huggingface.co/save_model/llama3-2k-non-ref-spa_2) on the datasets/training-llama3-2k-non-ref-spa_3 dataset. |
| | It achieves the following results on the evaluation set: |
| | - Loss: 1.5868 |
| | - Rewards/chosen: -19.2561 |
| | - Rewards/rejected: -19.7078 |
| | - Rewards/accuracies: 0.5294 |
| | - Rewards/margins: 0.4518 |
| | - Rewards/mix Margin: 0.4518 |
| | - Logps/rejected: -596.9339 |
| | - Logps/chosen: -612.0682 |
| | - Logits/rejected: -0.1396 |
| | - Logits/chosen: -0.2272 |
| |
|
| | ## Model description |
| |
|
| | More information needed |
| |
|
| | ## Intended uses & limitations |
| |
|
| | More information needed |
| |
|
| | ## Training and evaluation data |
| |
|
| | More information needed |
| |
|
| | ## Training procedure |
| |
|
| | ### Training hyperparameters |
| |
|
| | The following hyperparameters were used during training: |
| | - learning_rate: 5e-07 |
| | - train_batch_size: 1 |
| | - eval_batch_size: 1 |
| | - seed: 42 |
| | - distributed_type: multi-GPU |
| | - num_devices: 2 |
| | - gradient_accumulation_steps: 8 |
| | - total_train_batch_size: 16 |
| | - total_eval_batch_size: 2 |
| | - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
| | - lr_scheduler_type: cosine |
| | - lr_scheduler_warmup_ratio: 0.1 |
| | - num_epochs: 1 |
| | |
| | ### Training results |
| | |
| | | Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Rewards/mix Margin | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen | |
| | |:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:------------------:|:--------------:|:------------:|:---------------:|:-------------:| |
| | | 0.4129 | 0.53 | 1000 | 1.5868 | -19.2561 | -19.7078 | 0.5294 | 0.4518 | 0.4518 | -596.9339 | -612.0682 | -0.1396 | -0.2272 | |
| | |
| | |
| | ### Framework versions |
| | |
| | - Transformers 4.36.2 |
| | - Pytorch 2.1.2+cu121 |
| | - Datasets 2.14.6 |
| | - Tokenizers 0.15.2 |
| | |
